MovieStuff wrote:Paul, if you need assistance, call me and we can go over any problems you might be having.
My apologies, I just wanted to get more insight into what was going on before I waste your time on the phone seeking help with only a vague idea of the particulars of my issue. Will do a bit more testing when I get a chance (with some other long reels to ensure that it is indeed only happening specifically with that one) and then I will give you a ring! Many thanks!
MovieStuff wrote:Paul, if you need assistance, call me and we can go over any problems you might be having.
My apologies, I just wanted to get more insight into what was going on before I waste your time on the phone seeking help with only a vague idea of the particulars of my issue.
I appreciate that but, honestly, it is always best if customers just contact me directly. There is no one on any forum that is going to be guaranteed to provide you with 100% reliable information about my product and I am always available. So call when convenient and you are near the unit and I will be happy to help you.
1) Tscan, is there any chance you can upload your NeatVideo film profiles to the NeatVideo profile system, and/or provide a link here? I too use and LOVE NeatVideo.
2) One other tool that I have found useful for processing... This may be part of the "Video 9" system, which AFAIK is a packaged set of plugins/settings for VirtualDub/AviSynth:
-DeShaker, particularly the latest version which has some *radical* speed improvements for multicore CPU's. This not only can remove slight frame-to-frame misalignments, it can also clean up shake in the original shoot. Has some nice "magic tricks" including going back over as many frames as you dial in, to discover what was "outside" the current frame... and thus can literally smoothly reconstruct most sequences without having to crop anything at all. It also has automatic scene detection, which even the commercial solutions lack.
Great information Everyone. I'm ordering my Retro-8 today.
Few questions:
1. Does the machine store the original files of each frame in a folder, and do the sprockets show on these?
2. What is the image format,.. is it usable and viewable in its RAW mode? (ie. I take one frame to work with)
3. Can I pull all RAW image frame files into PremiereCC and make a movie there?
The images are stored in a proprietary flm format that can be exported as an image sequence or quicktime movie.
The image sequences can be exported as TIFF, JPEG, PNG, and BMP formats (you choose). They are exported into a folder under the name you choose for the capture. Sprocket holes are cropped off. The 1.33:1 images are pillarboxed (black bars on each side) on HD export, but not on SD export.
You can import any frame into a photo editor (e.g. Photoshop).
The image sequences can be assembled into a movie with any nonlinear editing program (e.g. Premiere, Vegas, et al).
Congratulations on your choice! It's a great little machine.
jrcasey1960 : Have any more good things thou hast learned? I'm an Adobe CC person. Love their stuff. So any insight will help and save me time I am certain.
Saw this post on another forum. Helps explain some things about the built in cropping..
My answer if the other 11% is needed, Is scan twice, high and low, and merge on Premiere or other editing software.
And yes, his statement of locking up the software is accurate. Just do it as he recommends and all will be well.
------------
...a couple more data points.
A little more digging into the raw capture file - I have my Settings checked to “Capture Uncompressed†in the RetroScanHD Pro software. The RetroScanHD software stores the film scan in a file with a .film extension. The data within the file is a series of BMP image files. The BMP image headers are easily searchable within the file and you can extract individual raw BMP images (I just used a hex editor). I extracted one BMP image from the capture file. The capture images are stored as 1002x720 pixel BMPs. I compared this with the exported 720p image and it was identical (with black pillar bars added to fill the sides). So it appears that the capture file already has ~11% upscaling applied before it saves the BMP images. I previously assumed it would have stored the unscaled image and performed scaling in post processing as part of the export function. I would have preferred to have had the raw capture unscaled to provide more accurate frame based processing outside of the RetroScan software. I’ve submitted this as a new feature request to the software developer.
I also discovered a pattern for the USB problem I was having that I mentioned in a previous post. On my system, I’m able to consistently crash the USB interface if I do the following order of operations:
1) Click Start in the RetroScanHD SW
2) Turn Switch On (Foward)
3) Click Stop in the RetroScanHD SW
4) Turn Switch Off (middle)
5) Click OK for the capture save
6) Click SCAN
7) Click Start
8) Turn Switch On (Forward) -> USB interface hangs
When I reverse steps 3 & 4 the interface never hangs. I’ve passed this info along to the software vendor.
The above may be my issue. Getting a new PC and will try it all from there. My current PC is good but highly loaded, but lacking some RAM.
Roger has made a great product. The software is very easy to work with after a short learning curve.
I will probably do post production in Adobe Premiere, taking out some grain and boosting other things that these modern programs are so good at.
Glad the unit is working well for you. Regarding software lock up, the correct order for using the software is to start the Retro unit running, then start the capture on the software. When finished, stop the software and then stop the Retro unit. Theoretically, it should not really matter what order you do this but some PCs are picky. So we recommend using this sequence.
Can someone provide a summary of the features available in the Retro-8 Pro that are not available in the original Retro-8? I am aware of some of them, including the speed difference, but thought a complete summary might be useful. This question would probably best be answered by someone that uses both versions of the machine.
Retro-8 speed = 2fps
Retro-8 Pro speed = 15fps switchable to 10fps for uncompressed captures
Retro-8 exposure = manual only
Retro-8 Pro exposure = manual or user defined auto, plus exposure boost, if needed
Retro-8 gamma = software slider controls gamma of capture frame
Retro-8 Pro gamma = software slider controls gamma of camera signal for cleaner handling of underexposed footage
Retro-8 registration = software based only
Retro-8 Pro registration = sensor or software based
Obviously, what most people are attracted to is the increased speed. The original Retro units required 4 hours to transfer 400 feet. That same roll can be transferred on the Retro Pro units in about 30 minutes. Also, because the original Retro units used software registration, you had to set the exposure so the sprocket hole was the brightest thing in view of the camera. This actually worked pretty well but, if you had really dark footage, it could be inconvenient. Also, if you had wide expanses of sky that intersected the sprocket holes and you had the exposure bumped up a bit too high, that could cause registration issues. Those rules still apply on the new unit if you have software registration (image stabilization) turned on but, if using the sensor for registering the frames, you can set your exposure anywhere you wish without affecting registration. This makes handling clear edge film a lot easier. The recently released 3.0.21 version of the software with the camera gamma control and exposure boost can work wonders on dealing with super black footage while minimizing grain.
Thanks for the concise summary. As you may know or have understood from my previous post, I am using an original Retro-8. Generally, it does a great job and so far I have not been too concerned about the speed. Most of the time I don’t have miles and miles of film on the shelf to be scanned, so I can work around the speed limitation by letting the unit run unattended and doing something else in parallel. However, the one aspect I do find troublesome is the registration. I find that scanning the same reel with somewhat different exposure (not necessarily significant difference), will give me different totals of captured frames, which of course means that some frames have been dropped in the one case. You gave me a complete explanation of this several months ago, so I fully understand what’s happening. It is caused by a bright area of the frame touching the sprocket hole. When this becomes really troublesome is when I run a large reel unattended and on return find that the scanning had stopped somewhere in the middle. This would be caused by too many frames being skipped and the unit assuming end of the reel. At that point what I have to do is to review how much footage had been captured, restart the scanning and at the noted footage be ready to manually turn the exposure down to avoid the dropped frames. Subsequently, the exposure needs to be returned to the previous level once the troublesome footage has passed. This can be at best very annoying and at worst quite time-consuming.
Based on your summary and explanation, I assume that the only way to get around the registration issue is to upgrade to a Pro unit. Is that still possible at this point?
