Crescendo Sys RTC2000 and gamma - Page 2

This isn't sounding like an upgrade

Anyone wanna buy my kimcoder MK1....

Kal,

The downside of a more agressive maximum gamma is the following: when the gamma is increased the ouput amplitude increases. In the case above, it goes up from 700mVpp to 800mVpp, there is nothing I can do about that. The more agressive the gamma the more the increase in amplitude. Some people can handle 1Vpp easily, while others already have an issue with 800mVpp. If a wider gamma control is desired I will give buyers the choice to set their maximum output voltage at the time of ordering.

For existing owners of a RTC2000 I will implement a true upgrade where you can buy a new board (after returning the old one) with gamma for just the price difference. It is an easy swap that anyone can perform when you can handle a screw driver.

Interesting Kim. Guess my problem (and possibly other people's problem) is that I don't know how much gamma I want. Been running SD-DVD off of an HTPC for almost 6 years now and I use 2-3 notches of gamma boost in the ATI control panel (running overlay). No idea how that actually would look if graphed.

IMHO, if you need to choose, I think most people would aim fairly low in gamma boost. Anything too high looks like crap.

Kal

Tse,

If you are around, maybe you can give some input on a decent maximum gamma curve to shoot for.

Kim

Kal,

You can choose the gamma you want, that's not the problem. The problem I see is that it changes the output amplitude and that may be an issue for some people.


KIm

Old but interesting article on gamma and CRT's ...

http://scien.stanford.edu/jfsite/Pap...rrell_1989.pdf

Sorry, yes, that's what I understood. You're just trying to figure out what sort of range to offer since if you offer too large of a range (ie: allowing a steeper curve) then the output amplitude may be problem. My suggestion was to not offer any sort of crazy high gamma boost if that causes problems. Gamma should be used sparingly but most people new to the concept will tend to use "more than less" at first.

You'll figure out a good curve, I'm sure.

Kal

We call 'em kimcoders.

You guys crack me up. I'll tell you what. How about I get a kimcoder standard and one custom with 5 BNC. On the standard one, I'll use a very short VGA to 5 F BNC breakout. Then both will use a 25' run of belden 1694a with Canare connectors.

Well do a single blind test. 10 runs. No more than 6 of one (may be 5 and 5 or 4 and 6, you won't know). Each run, you wirte down whether you thought it was the VGA with short breakout or the BNC one.

I'll bet you any amount you want that you can't get it 100% correct.

Dave

Wow!
That is great, I honestly didn't expect there to be an upgrade path.

Umm how is it you Americans put it?


*ahem*


YOU ROCK!



yes that's it.

All,

I just send out the second set of test boards for the transcoder with gamma. I should get them back in about a week. I will keep you informed.


Kim

Thanks Kim!

Kal

I tested different settings at on tse´s desing. When i changed R1 ( 2.4k) to smaller value(1.5k), to make more agressive gamma curve then output goes over 0.7vpp. if i remember correctly, then to get output back to 0.7vpp R2 was over 1k (1.2-1.5K)
So if you want to make adjustable gamma settings you must ajust those two values together, maybe set of resistors and jumpers to change?

Just my 2 cents but from what I was told from one the testers of Kim's first version of his new transcoder with gamma, the amount of gamma was more then adequate. If people feel they need more or want to play with the curve then they could do so themselves by changing a resistor or two (pretty easy) based on input from other users.

kal

The breakpoint in the gamma curve can be continuously changed between 0 IRE (no gamma) and 30 IRE (maximum gamma). At this time that control range seems to be more than enough. If it appears it has to be changed later that can easily be done.


Kim

That should be plenty.
Any idea at all on a release date?

When it works I can start production real soon.


Kim

I just finished putting one of the test boards together and I did some testing. So far it looks good, the blacklevel clamp seems to work as expected and it doesn't shift with changeing picture content.

I will keep you all informed on further measurements.


Kim

Are you going to post "upgrade" instructions here, or you website, or both ? GREAT idea, btw, I've seen your transcoder run with/without a Tse gamma board, and it definitely looks better WITH the Tse gamma board in the circuit.
Can't wait !

When I have the confirmation that it all works as intended and ordered the boards I will put the upgrade path on my website.

Kim

Can't wait!

I think I'll need to upgrade also. Just got my RTC 2200 that I got from you right before Thanksgiving setup. Output from DTV HR20-700 looks pretty good. Output from Playstation and XBox seems to have crush. Would this fix?

If the crush is in the source--most likely no. If the crush is in your PJ--yes.

Dave

I guess to check to see if its source versus projector would be to throw up an internal test pattern? I'll have to check to see if my Marquee 8000 has this, but I think it does.

The only thing that confuses me then, would be why a gamma correction in the signal path BEFORE the projector wouldn't correct it either way....never mind, I think I just thought it out replying to this....so if the SOURCE crushes the black levels, there isn't a way to get them back. But if the SOURCE provides proper black levels, the projector is crushing them because it isn't seeing enough variance in the levels internally to differentiate. The gamma correction will "stretch" out the black levels to give them more "range" (sorry for my newbie wording) so the projector has more to work with. Is that sounds correct?

Well, it depends upon how it is crushing them, but yes. If it sends out a 2 IRE signal that is basically the same as a 1 IRE signal, there is no way to get it back.


You pretty much hit it on the head. With a CRT, the phosphor does not react enough such that you can have a pitch black 0 IRE and still have visible light at like 2 IRE. So, folks with CRTs decide to have a full fade to black and crush blacks, or have elevated absolute black (but still darker than digitals) with good shadow detail, or split the difference. These types of gamma circuits are sort of a "best of both worlds" approach.

Dave

I think that Dave hit the nail on the head. The crush that a CRT performs is a case of "smooth" crush in which the response is diminished from the ideal.

Depending on the engineering qualities of your source (Xbox in your case) when the crush is not smooth but anything below a certain level is just plain cut off than any amount of gamma won't do anything. In the end zero times something is still zero.


Kim

Well then, HOW IN THE HELL CAN I BE IN XBOX NIRVANA!!!???

So, it sounds like I may be SOL unless the PROJECTOR is doing the crush. I'll try to check tonight. Hopefully that's the case. If so, Kim, when would I be able to get my hands ona gamma correcting box?

Thanks for you guys' help. I really appreciate your patience with a technical newbie.

Dave,

Do you think I need to make any more changes to the transcoder with gamma or is it ready for production? When I still want boards before Xmas I should order real soon.


Kim

So what do ya think Dave ?
Tom.W