Palacono's Bugs'n'Misunderstandings [#52 Hitfilm Framerate issues (Link only)]

New Bug?: #51 Grading Transfer in HFP2017 Not as good as HF3P is further down this thread
-------------------------------------   
New Misunderstanding?: #50 Proxy/Export Overrides Particle Preview Flag is further down this thread
-------------------------------------  

New Feature?: #49 "Real" 3D Text in all Hitfilms is further down this thread
------------------------------------- 
New Feature?: #48 Matte Cleaner Choke Texture Density is further down this thread.
-------------------------------------
New Bug:#47 Lens Blur Error in Manual and Scaling Bug? is further down this thread.
-------------------------------------
New Bug: #46 Hitfilm Doesn't like this 'Woodcut' Font is further down this thread.
-------------------------------------
New Bug: #45 Set Matte Enthusiasm Bug, HF1 to HFP2017 is further down this thread.
-------------------------------------
New Bug: #44 Loose Tracking Bug, HF3P to HFP2017 is further down this thread.
-------------------------------------
New Bug: #43 Ripple Edit Annoyance HFP2017 HF4P/E is further down this thread.
-------------------------------------
New Bug/Feature: #42 HFP2017 HF4P HF4E) Transition Layer Confusion is further down this thread.
-------------------------------------
New Bug/Feature: #41 HF4E Proxy File Deletion Bug is further down this thread.
 -------------------------------------
New Bug/Feature: #40 HF4 HFP2017 Grade Layer Gremlins is further down this thread.
 -------------------------------------
New Bug/Feature: #39 HFP2017 Opacity Conundrum is further down this thread.
-------------------------------------
New Bug/Feature: #38 HF4E HF4P HFP2017 Difference Key Issues
is further down this thread.
-------------------------------------

New Bug/Feature: #37 HFP2017 Proxy Bug #5 + Editor "Features" is further down this thread.
-------------------------------------

New Bug/Feature: #36 HFP2017 Neon Path Mask Offset?
is further down this thread.
-------------------------------------

New Bugs: #35 HP4 RAM Preview, Speed, Proxy Bugs
is further down this thread.
-------------------------------------

New Whine: #34 HP4 .MP4 file Render Quality is further down this thread.
-------------------------------------

New Bug/feature: #33 HF4 Parenting Polar Pothole
is further down this thread.
-------------------------------------

New Bug/feature: #32 HF4 Tracking Semi Keyframes
is further down this thread.
-------------------------------------

New Bug/feature: #31 Export Speed Extra Digits
 is further down this thread.
-------------------------------------
New Bug/feature: #30 "Invisible" Keyframes Bug is further down this thread.
-------------------------------------
New Bug/feature: #29 Blurry Patterns is further down this thread.
-------------------------------------
New Bug/feature: #28 Alpha Weirdness/Bug is further down this thread.
-------------------------------------
New Bug/feature: #27 Towards Layer Weirdness is further down this thread.
-------------------------------------

New Bug/feature: #26 Slow Delete is further down this thread.
-------------------------------------
New Bug/feature: #25 Dependency Bug? is further down this thread.
-------------------------------------
New Bug/feature: #24 Set Matte Weirdness is further down this thread.
-------------------------------------

New Bug/feature: #23 Set Matte Setback? is further down this thread.
-------------------------------------

New Bug/feature: #22 Linked for Life? Bug is further down this thread.
-------------------------------------
New Bug/feature: #21 Motion Blur/Motion Jump Bug is further down this thread.
-------------------------------------
New Bug/feature: #20 Quad Warp/Motion Blur Bug is further down this page.
-------------------------------------
New Whine: #19 Fiddly 3D Object Selection is further down this page
-------------------------------------
New Bug/feature: #18 Quad Warp Problem Deja Vu is further down this page.
-------------------------------------
New Bug/feature: #17 Magnify Effect Bug/Feature is further down this page.
-------------------------------------
New Bug/feature: #16 Text Outline Bug/Feature is further down this page.
-------------------------------------

New Bug/feature: #15 Viewer Playback Problems is further down this page.
-------------------------------------

Suggestion: #14 Rate Stretch Indication is further down this page.
-------------------------------------
New Bug/feature: #13 Text Layer Limitations? is further down this page.
-------------------------------------
Just me moaning: #12 Various Fixes? is further down this page.
-------------------------------------
New Bug/feature: #11 Codec Fun 'n' Games is further down this page.
-------------------------------------
New Bug/feature: #10 Grade Layers Problem is further down this page.
-------------------------------------
New Bug/feature: #9 Projector Resolution is further down this page.
-------------------------------------
#8 Proxy Edit Problems
It's hard for me to keep track of the bugs (or just misunderstandings of the inner workings of Hitfilm) I've reported, so I thought I'd start a thread and add them here and update the title occasionally. I'll probably link my other threads here later, so people can be aware and work around them until fixed in any future updates - although I do seem to find some freaky edge cases, so presumably not many people are affected by them... although everyone should be by this one.

Latest candidate is Editing with Proxied files.

The short version is that with HF3E or HF3P you can't if you want to be mildly creative.

The long version is: you can edit with Proxy files in HF3E and HF3P but your Transitions won't work properly, so you can't really use any.
They work in HF4P, but it takes up to twice as long to create a proxy file -  on a 5 second long 1280x720 test image.

IMO, this needs a patch, even though HF4P is out, because basic functionality is missing. Although: is  no one actually using the Editor for editing (or just not using transitions?), for this not to have been spotted before now?

The video's 4 minutes long, so if you don't want to invest that amount of time, the problem looks like this when you Transition between two clips. The Black section is supposed to be the other image:

http://i854.photobucket.com/albums/ab106/pickaname2/ProxyImage_3.jpg

Curious? Watch the video. ;)

https://www.youtube.com/watch?v=AS2LxxPAoaY

Some extra info: Both HF3P and HF4P were in 16 bit float mode and  HF4P takes twice as long to create a proxy file as HF3P : 32 seconds vs 16 seconds.

Tested in 8 bit mode, HF4P is only slightly slower: 19 seconds vs 16 seconds from HF3P again.

I was hoping HF3P would be even faster in 8 bit mode - it certainly is when editing - but maybe my HDD/oldish PC is the bottleneck. YMMV. :)

So, if it's at all possible to speed up HF4P proxies (and still work properly) to the HF3P speeds - after all it wasn't being stressed on a single image - as well as patch HF3P and HF3E, I'm sure that would be appreciated.

Not videod, but the same proxy Transition issues are also in HF3E, so that needs a patch too for anyone going the 'Buying Addons' path.

Tagging the team as usual: @Ady , @SimonKJones (and @Triem23 because he uses proxy a lot... ;) )

«13456710

Comments

  • @Palacono I just tried this in 3 Pro and Express on one system and 4 Pro on another. I was not able to recreate your initial results. In other words everything was fine once the first composite shot proxy was made where as you experienced a problem right away. Also my proxy creation times averaged 4 seconds or less across all three versions. For reference I was using 8 megapixel stills.

    Now I was able to recreate your other results by removing the transition and adding a different one after the proxy was made however I think this is normal. Creating the proxy is essentially the same as rendering out the timeline. Just like with rendering, changing a transition after the fact means having to render again to see the new result. Put another way creating a proxy after a transition has been applied and removing or changing the transition invalidates the proxy. When I was trying this canceling the existing proxy and creating a new one with the new transition applied cured any issues. The better way to go is create the proxy first then apply any transitions. 

    None of that explains why you had a problem right after creating a proxy though or why it's not happening to me as well.

  • edited February 2016

    @Aladdin4d, well some consensus is at least interesting.  8 Mega pixel stills are a different variable, so perhaps try again (thanks for trying in the first place, BTW) with a 1280x720 image and see if the behavior changes?  You've also clearly got a faster PC than me; which wouldn't be difficult. :)

    I'd initially tried this with a 1080p composite that itself contained an embedded composite (which was really slow) and when that didn't work, I thought I'd just lower the bar as much as possible to not over-complicate things.

    Hmmm... adding a transition and then making a proxy causes problems too? Not tried that, although I will after I've posted this. I can see how removing the effect might invalidate the proxy there - although that's a different kind of proxy, not even thought of how that should work or why you'd want to. The transitions aren't very intensive, but for a really slow PC: maybe?

    But even if it invalidates the proxy-with-a-transition-on-it (if it is different), it should still just work normally when you remove the transition, but be slower, just like on the composite timeline. Otherwise that would be like invalidating a proxy in a composite causing it to go blank, which no one's reported as that would be spotted pretty quickly. Nope, see below.

  • edited February 2016

    @Aladdin4d OK, tried to create the Proxy after adding the Iris Transition and it remained broken. In fact if the playhead is in the right place and the right file is proxied, the viewer window will change from semi-transitioned to semi-plus-a-missing-bit as soon as the proxy process completes.

    And if I proxy both images with the transition already applied, it breaks in exactly the same way as if I'd proxied both of them before applying the transition.

    So, creating the proxies either before or after adding the transition makes no difference; it breaks in the same way. There is only one type of proxy and the transitions are just added on top. Which makes sense really, otherwise you'd have Transitions visible back in the Composite timeline, or need a second proxy for that.

    And to be completely clear in my own mind: :) I proxied both files before adding any transition at all, then added Iris and then Push, and it was broken in the usual way; so order isn't important.

  • Aladdin4dAladdin4d Moderator
    edited February 2016

    @Palacono ;Ignore everything I said earlier I should've paid more attention.

    Ok so now I can duplicate what's happening but only with transitions that require Clip A to be extended beyond its original length.

    What's happening is a proxy isn't allowing itself to be extended to complete a transition the same way non-proxy media is so when you apply the right kind of transition everything is fine until you reach the end of Clip A. Normally Clip A is artificially extended to the end of the transition but a proxy just drops out completely.

    The reverse is also true for Clip B - If it has to be extended before it originally begins it drops out too.  

     

  • edited February 2016

    @Aladdin4d Good point and that is correct for one side of the transition, but not the other.

    I'd forgotten I'd tried increasingly further cuts into the original 1080p video comp that started this investigation and couldn't get it to work; but here's a video showing even with loads of proxy footage available, it still isn't using it all.

    https://www.youtube.com/watch?v=IXzocicc1BE

    Your suggestion that it is not extending the clip beyond the end of the proxy - but is with the original - is an interesting one. With an image you wouldn't notice if the frame was held (although it doesn't know that it is an image) so what's it going to do doing with video (in HF4P, it does nothing in HF3P)? Holding the last frame or using video frames beyond the cut?

    As it doesn't work with video or images in HF3P, I've not yet tried it with HF4P to see if it suddenly freezes during the transition, which could cause problems in itself, depending on how it's implemented. What if you have a 4 second transition and only 1 second of video available after the cut (2 seconds per side)? Does it show the one second available on one side then hold the last frame of that for the next second, or just hold the frame at the cut point and not use that remaining second?

    I guess I'm going to have to try it with a section of video with a timer on it and see what it actually does. :) I really do not want it to freeze during transitions, so hope it doesn't.

  • Guys @Palacono, @Allandin4d I am interested to know, does the problem tranfer from the proxy to the export, or is Hitfilm using a fresh export render which then works?

  • edited March 2016

    @Andy001z sorry for delay, meant to check this out yesterday.

    If you haven't yourself: yes, the exported file is broken in exactly the same way as in the Editor in HF3E and HF3P when using proxy files. So it's completely WYSIWYG in that respect.

    (And you spelt @Aladdin4d 's name wrong. It's OK, I keep spelling it with 2 ls myself. ;))

  • Triem23Triem23 Moderator

    Quick note on proxies @Andy001z ;

    By default proxies are used for export. This can be turned off in the Hitfilm's option menu. 

    My guess is if you don't have enough tail on a clip to cover a transition, outgoing clip just pops away. 

  • edited March 2016

    @Triem23 nope, it's a bug, watch the second movie. :) In HF4P it's been fixed and works completely as expected (although proxies are a lot slower to produce. Does this also effect render times in general? Not checked that, as I'm using the HF4P demo). I did a check with timecodes on the video in the Editor and a second on the composites. They ran differently with one stopping and the other not in HF3E and HF3P, but not in HF4P.

    In HF4P they work as they should. Any unused head or tail left (with or without a proxy file) is used in the transition section. Anything beyond that (either end) is replaced by the first, or last, frame being held to stretch to the length of the transition.

    So, if you put a cut at one second from the end of a clip, but then use a 2 second transition on one side (4 seconds total), it will show the first second, then hold the last frame for the last second to make up the 2 seconds.

    Frustrating if you've already waited ages for a proxy to be made, then have to be careful with your transitions (basically you can't have any). The alternative is to invalidate the proxy and have it render the lot from scratch, which obviously takes a lot longer.

    Plus, I don't understand why proxies take twice as long to produce in HF4P as HF3P. Same PC, same folder, same project file, same everything. So one step forward, one step backwards.

  • Triem23Triem23 Moderator

    It might depend on what you're rendering. In general, I assume anything using 3D (especially models) would be slower in 4 because the new render engine--physical shading and reflections--uses more overhead than the previous engine. For 2D/3D planar stuff I can also see a bit of slowdown if using manual bezier curves, but I speculate. 

    I am surprised Hitfilm holds the last frame to extend the transition. Vegas, Premiere and Avid don't. 

    Proxies. I have a strange relationship with proxies--I tend to only use a proxy on "sub clips..." so, actual example: creating a scene with multiple 4K planes of Caustics, I'll proxy out the fractal noise sources for the height map. Near-final scenes, however, I will render out as an image sequence to re-import. The aforementioned Caustics shot is an example--it took 6 hours to proxy, during which time editing was sluggish. The render took a tick under four hours. And the image sequence takes less drive space, can't be accidentally invalidated, and isn't automatically cleared in 30 days. The render is enough for timing and trimming purposes. For this particular shot I needed to add a distortion layer, but it was actually faster to set up and mask the distortion over the render (to render again) than it would have been to add the effect in the original comp and reproxy/render that way. 

    But it depends on the project I guess. 

  •  I can't seem to replicate this. Can you provide a simple step-by-step on how to recreate?

  • Aladdin4dAladdin4d Moderator

    @SimonKJones The first video shows how to recreate it step by step pretty well just be sure to use the same type of transition i.e. one that shows both clips for the duration like Iris. To recap place two stills on the edit timeline, comp one (or both) and make a proxy then apply a transition like Iris. Just a repost of the first video

    https://www.youtube.com/watch?v=AS2LxxPAoaY

     

  • edited March 2016

    @SimonKJones . @Triem23 the first video shows all the steps from the beginning to the end with two 1280x720 images.

    Edit: Ah, @Aladdin4d beat me while I was writing this post and waiting for the video to upload. ;)

    This is for HF3E and HF3P only. HF4P has fixed it. Would be nice to have a patch for those others though. And the new Quad Warp? ;)

    For a complete comparison of HF3E (HF3P is the same; see first video ) and HF4P, including how they extend the first/last frames when they run out of of proxy (or unproxied, actually) video, this should answer any questions.

    Note: It's 7 minutes of me playing with every example I can think of, so not exactly an action video. :)

    https://www.youtube.com/watch?v=R4YzR07-Dvw

  • FWIW, I also ran into problems with proxies and transitions in my very first hour of using HitFilm.  Noticed in 3, seemed fixed in 4 Pro.

  • edited March 2016

    OK, new bug, or feature. #9 Projector Effect Resolution 

    Tagging @SimonKJones as he's the expert on Projector.

    I'd assumed that Projection worked 'through' the plane it was on and didn't care what the scale was, but apparently it does, which gave me some odd results. I first noticed that the projected image looked pretty soft, so investigated why that might be.

    See video - with almost everything that could confuse the issue removed, including a moving camera that was tracked in mocha.

    I suppose I could make a projection plane much larger than I need, then scale it down again, but that seems a bit inefficient and Hitfilm's probably having to do more unnecessary work.

    Plus...why? :)

    Oh, to be a bit clearer: Using a Projection at a scale of 1:1 back onto the original image is blurrier than the original. Changing the resolution of the plane being used has interesting effects on that, but even when the pixel matrix is higher than the original was, it's still blurrier. Add in the use of HF4P's Quad Warp which now has Anti Aliasing applied (not shown in this video, it's still pretty soft with HF3P's version) then it's softer still.

    https://www.youtube.com/watch?v=lyxfQQd4fvc

  • Triem23Triem23 Moderator

    When defining a plane, one is defining a pixel matrix which the projector effect is mapping pixels to. As you've noticed the resolution of the screen plane will affect the quality of the projected image precisely because the initial size  of the plane determines an absolute number of dots. This is mathematically similar to how reflections and shadow maps work, which is why those controls were added in HF4. 

    All of my projection setups are done using large planes, but part of that setting up anchor points on a bottom edge of my ground plane, then duplicating that plane for other screens. Only one plane in the media pool, and the Grid and Projector effects are copied to the duplicate. 

  • edited March 2016

    @Triem23 Sure, I know it's doing something, but why was the original Projection, which was at 100%, still softer than the thing it was projecting?  There should have been nothing done with the pixel matrix other than "Copy pixels @ 1:1. Finiiiiiished! ;)"

    Edit: Actually, as I was scaling the texture down before I zoomed in, I was increasing the density of the Pixel Matrix and it was still soft.

    In the later portion when it was smaller: what useful benefit is there for it to take the projection, scale it down to the original size of the plane (proxying that result somewhere...), then scaling it back up again? Skip a step (or three) and everyone's happy. :)

  • edited March 2016

    New bug, or feature. #10 Grade Layer Problems

    I know there are order of operations things that can account for some of the behavior I'm seeing, but I think this is more than that.

    When putting one Grade Layer on something doesn't work, then two will be better, right? Or at the very least, do nothing bad? Well, apparently not.

    Already having some slight struggles with Grade Layers in this thread Masking Grade Behaviour  some of which may be (probably is) due to known (to FXHome anyway) Order of Operations features, but seems weird that I can't use a Mask on a Grade Layer. Unless I need to put another one above it. But I tried that and it didn't work.

    However, something different happened in this video, where Grade Layers + Grade Layer = 0 Layers... or sometimes one. For a single frame.  

    Actually in general, Projector (which I think might be partly responsible) often does not update the viewer window when a parameter is changed until the timeline is moved a little.

    This is the same in HF3P and the HF4P Demo - I've been using HF3P, but opened the project in HFP4 to check, and made the video of that.

    Plus two 100% Sharpens = Yikes! I can massively over-sharpen something in a photo package and it doesn't ever look like that. It gets the usual 'halo' effect around things, but this just pushed the colours right off the scale and back around again.

    So, @SimonKJones , @Ady ? Over to you. :)

    https://www.youtube.com/watch?v=ky7MJM8naIQ

  • edited March 2016

    New bug, or feature. #11 Codec Fun 'n' Games :D

    It's often been mentioned that Hitfilm struggles with .MP4 /H.264 files, and especially when tracking. It regularly becomes unresponsive and tracks until it gets to the end or goes out of range, whether you want it to or not. For that reason I usually do long or complicated tracks in Hitfilm 2 Express then save the results and import them into Hitfilm 3, because that stays responsive throughout and is also a lot quicker anyway with the same video files.

    It's mostly suggested that you should just use a different file format and all the problems will go away. Well that doesn't work, because it also locks me out when tracking Cineform .AVI files, which are specifically designed to be 'Editor Friendly'. It's a case of diminishing returns anyway, as the read speed of your HDD starts to become an issue with bigger files.

    Hitfilm seems to get slower with each new version. Proxy files take twice as long in HF4P as they do in HF3P even for a single image, as there is (I assume) more and more general overhead.

    But what about .MP4 files for general editing?

    Well, they are apparently the Spawn of the Devil and should be avoided, but it wasn't always the case. This video shows a problem file in Hitfilm 4 (and 3) that works perfectly well (and pretty fast too, no judders or 'catching up' ) in Hitfilm 2 Express (end section of video).

    Although this is a slightly weird .MP43 file in an .AVI wrapper, the point is still the same: something changed for the worse from HF3P onwards.

    https://www.youtube.com/watch?v=wBehRslLN1g

    So, @SimonKJones, @Ady , @AxelWilkinson can we please, please have the old .MP4 handling code back in a future update for Hitfilm and stop telling people they need a new graphics card, hard drive, processor and have to convert everything to DNxHD to get a decent speed out of Hitfilm?

    And while someone is in there, perhaps they see why proxy files are so much slower to create (twice as long) in HF4P than HF3P, even for a simple, single image displaying for 5 seconds and nothing else? see Proxy Problems video at the top of this thread for the comparison.

  • Aladdin4dAladdin4d Moderator
    edited March 2016

    @Palacono I totally get where you're going with #11 Codec Fun 'n' Games but an MP43 encoded clip is a really bad choice for comparison. 

    First MP43 is MS-MPEG4 Version 3 which sounds like it would be a common MP4/264 type clip except it isn't. It's a distant redneck cousin nobody likes to talk about and when they do there's rumors of inbreeding because that cousin, well "he just ain't right." MP43 is closer to Divx and Xvid than it is true 264. All three are an attempt to use a very small subset of the 264 spec in a way that's fully Video for Windows compliant and a lot was garbled in the translation

    Second decoding MP43, or any AVI for that matter, won't use the same pipeline as true 264 compliant clips. HitFilm uses Video for Windows to handle AVI files which is old and deprecated many times over but is still heavily used by most if not all Windows NLE's. There are actually several good reasons for that but the biggest ones are DirectShow is a pain to program around and its decoding usually isn't frame accurate and it's an easy way around some licensing issues in the same way using the QuickTime framework gets you around licensing issues.

    Third HitFilm uses the MainConcept framework so if you really want to test MP4 performance across versions and configurations it needs to be an MP4 file. Anything AVI is going to use VfW and most MOV's will use QuickTime. I say most for MOV's because HitFilm 3 and up appear to ship with a native DNxHD/HR decoding dll so presumably QuickTime gets bypassed for DNx decoding.

  • edited March 2016

    @Aladdin4d , I'm sure all that's perfectly correct, but at the end of the day: Hitfilm's 3 and 4  don't work as well with any type of (proper) .MP4 format as Hitfilm 2 Express did/does. The tracking issue shows it up and it's just generally slower. If you still have access to HF2E (or quite probably HF2U) you should be able to see the difference yourself.

    It was just that that particular inbred cousin actually also showed that the later Hitfilms don't even handle it visually consistently in two different places in the same program as well as slowly.

    Sadly, something got a "new improved recipe for your convenience" that's just not as good as it was.

    Library wrappers on top of wrappers on top of indirection on top of other stuff is probably to blame. It's like drilling down through some early Windows code to find a simple DOS call at the end of the chain. :)

  • Aladdin4dAladdin4d Moderator

    @Palacono I'm not disagreeing with any of this:

    "I'm sure all that's perfectly correct, but at the end of the day: Hitfilm's 3 and 4  don't work as well with any type of (proper) .MP4 format as Hitfilm 2 Express did/does. The tracking issue shows it up and it's just generally slower. If you still have access to HF2E (or quite probably HF2U) you should be able to see the difference yourself."

    Everything I said was just to point out how MP43 doesn't have anything to do with MP4 decoding which I thought was what you were trying to focus on. 

  • @Aladdin4d, OK, just disappointed with decoding in general, really. That one just had a visual payload. :)

  • @Palacono some interesting comments and demonstrations, I too noticed a drop in performance moving from Hitfilm 2 Express to Hitfilm 3 Express or Hitfilm 4 Pro. I just put it down to the software under the hood having to do more and therefore struggling more on my aging PC architecture.

  • Aladdin4d

      It's a distant redneck cousin nobody likes to talk about and when they do there's rumors if inbreeding because that cousin, well "he just ain't right."

    You owe me a keyboard cleaning, I spewed coffee all over it laughing at that comment. :P   Nice one

  • Aladdin4dAladdin4d Moderator
    edited March 2016

    @FlyingBanana78 Do I look like McDonald's? I cannot be held liable for your negligent handling of a common hot breakfast beverage of unknown origin or any damage that might result from the negligent handling of said hot breakfast beverage. It's hot coffee. Drink responsibly ;) 

  • edited March 2016

    @Aladdin4d So why do you wear the clown makeup, red wig and oversized shoes?   :)

  • Aladdin4dAladdin4d Moderator
    edited March 2016

    @Palacono Oh this I'm wearing right now? This is just what I wore to a pitch meeting for a Killer Klowns from Outer Space TV show  this morning and I just haven't had time to change yet.  It's going to be a trilogy in four parts.

  • New Bug/feature: #12 Various Fixes?
    -------------------------------------
    Not really a bug, more a complaint. Whatever 'Various Fixes' were in Update #11 for Hitfilm 3 Pro, they did not include fixing the Proxy Transitions bug, so they're still unusable. and (I was hopeful they'd copy the code over, but seems not) Quad Warp is still wobbly as heck, and therefore: also pretty unusable; especially with tracked points.

    Well, I'm not going to upgrade to HF4P to get bug fixes.

    If you want the extra features in HF4P, by all means: upgrade. But Bug fixes? They should be corrected in updates. If the new features are good enough, people will still upgrade to HF4P, so you're not cannibalising sales. Dumping previous customers is bad form, especially when they've already found and fixed the bugs in HF4P. What's the real reason for not copying the functions over to HF3P, recompiling and releasing an  update with the bugs fixed? Money?

    New real bug coming soon. Text does not work like other planes, which limits its usefulness.

  • @Palacono do you work for the firm that shall not be named ?

Sign in to comment

Leave a Comment