Pues sí tengo sonda. La spider 5 pro. Ahí la tengo para cuando tenga la paciencia de ponerme a fondo. Pero me desilusiona que no sea práctica para calibrar HDR.
Pues sí tengo sonda. La spider 5 pro. Ahí la tengo para cuando tenga la paciencia de ponerme a fondo. Pero me desilusiona que no sea práctica para calibrar HDR.
Última edición por MACH 5150; 20/02/2017 a las 10:09
Yo no creo que no sea práctica, sobre todo, si ya la tienes. Tal vez no te dé un resultado super exacto como una de esas que cuestan un cojón de pato o una spectracal C6 con certificación HDR que tampoco es barata pero EMHO siempre será mucho mejor que ninguna calibración. En principio, para mirar la luminosidad poco tienes que hacer. Poner el iris en abierto total un patrón 100 IRE y listo. Después, para calcular el brillo objetivo lo multiplicas por 6 o por 5,5 si lo quieres con un poco más de brillo y ya tienes todos los datos para generar tu propia gamma.
Venga ánimo que me siento un poco tonto hablando de las bondades de esto yo sólo.![]()
Hola. Esto hay que hacerlo considerando que previamente hemos seleccionado el Color Profile BT2020, ya sea el "oficial" de JVC o algunos de los existentes en AVSForum (con y sin filtro). La diferencia de luz medida en función del Profile seleccionado puede ser muy significativa.
Sldos.
Proyectores: JVC DLA-X7000B, Infocus IN78, Infocus SP4805
Pantalla: Da-Lite Contour Electrol 106" (gan. 1,1)
Fuentes: Chinoppo M9702 V2, HTPC, Server HP Proliant
Receptor AV: Yamaha RX-A3030
Cajas: 2-Keft Q55.2, 1-Keft Cresta C, 2-Keft Cresta 2, 2-B&W 686, 1-Subwoofer Rel QUAKE
Tienes razón, lo daba por supuesto. Lo que pasa es que el objetivo de medir la luminancia, en este caso, es poder seleccionar los parámetros para generar la gamma y se supone que vas a ver las pelis en las mismas circunstancias con las que has medido con lo cual deja de ser irrelevante si era o no era sin filtro porque aunque la luminancia que obtienes es diferente, también lo es la gamma que obtienes y por lo tanto coherente con cada una de las circunstancias.
Supongamos que usando BT2020 con filtro obtienes en alta 150 nits. Esto implica que tu objetivo de brillo sería alrededor de 900 nits. Si usamos BT2020 sin filtro tendrías unos 165 nits y tu objetivo de brillo sería alrededor de 1000 nits. Por lo tanto, las dos gammas serían diferentes y coherentes con cada una de las circunstancias de visualización.
Última edición por Anonimo08012019; 20/02/2017 a las 11:07
De momento mi intención es probar las gammas que manni tiene colgadas, porque en principio es sencillo y da una buena idea de las posibilidades. Estos días no tengo más ganas de entrar en profundidad en el tema de la autocalibración.
Pero necesito una clarificación previa que no acabo de entender.
El manni dice que ahora las curvas buenas son las generadas con el soft casero. Este es el párrafo en cuestión:
I use a new file name standard, with JVC in the front to make the difference with the old files, which can all go to the bin. There is no need for the old first parameter as there is no calibration involved, it's all mathematical. My PJ isn't part of the process at all.
The new files give the target, the beginning of the roll-off and the clipping point. So the first parameter is like the second in the old file name convention, the second one is new and the last one is the same as before. [Edit: this might change as I clarify the way to use the target/max brightness parameter].
For example, my last curve was 200-1100-4000. The new version here is called JVC-1100-800-4000, because I wanted to start rolling off from 800nits. This is a much better version of what I was attempting to do with Calman and the JVC Autocal.
https://www.dropbox.com/s/l5qy28hmqg...-4000.jgd?dl=0
Then I also quickly generated a JVC-1000-240-4000, which starts rolling off at 240nits and should be a bit brighter, but with a much softer roll-off:
https://www.dropbox.com/s/eyomn6g0l9...-4000.jgd?dl=0
A JVC-800-500-4000 which should be a brighter for larger screens
https://www.dropbox.com/s/tazufv5uzk...-4000.jgd?dl=0
And a JVC-600-400-4000 for super large / low gain screens wich should be significantly brighter:
https://www.dropbox.com/s/v2givniq3s...-4000.jgd?dl=0
This should give a decent starting point for many people, other than this create your own curve or ask your calibrator to do so. Please don't ask me how the tool work, others will soon explain and share more curves.
I haven't had the time to do any extensive testing, and I could have got some of the parameters wrong (I still need to figure out how max brightness exactly works), but overall they should work well.
Mi duda es si las puedo importar con autocal o hay que usar necesariamente el soft rupestre ese que no tiene ni menús.
Lo digo porque el manni dice que con el autocal se pierde toda la gracia:
Originally Posted by sonichartI feel like a dope and I'm probably just a little hungover from last night, but I was originally under the impression that the new JVC curves were created with Arves tool and are, thus, better.
If I upload those curves with JVC Autocal am I losing out on the goodness in those curves?
I just started learning Python ~6 months ago, so kinda excited to get to use the tool if need be
The new curves created with Arve's tool are marginally better if you upload them directly to the JVC because they use the full internal resolution of the JVC (256 points vs only 11 when using the JVC Autocal gamma feature or import/export feature). I exported these new curves to give a low res (11 points) for people to compare, and I also created a few curves with a lower multiplier for people who needed more brightness, as with the tool it takes seconds instead of hours when done manually.
So you're losing most of the goodness if you import the JVC curve that is the equivalent of my most recent manual curves, but if you need brighter curves then the new JVC ones should make a significant difference.
Overall, Arve's tool offers speed and a lot more flexibility to shape the curve, but the added resolution will be more significant in some cases than in others. I have noticed that the full resolution one gives nicer gradation in the highlights, but you really have to be pixel peeping to notice it.
I'm not going to write this again because it's all in the thread and I will spend my time repeating myself.
If anyone asks a question and I don't answer, it's because I've already answered it or because the info is already in the first post of the JVC autocal thread or in one of the posts linked there in the HDR/Gamma section.
I've also said that would not answer any question about using Arve's tool or installing it, I don't have the time to do that but others will I'm sure.
Lots of advanced users have got it now so they'll take over and will probably share some great curves soon.
Entonces, ¿tiene algún sentido meter las custom gammas nuevas con el autocal o es una pérdida de tiempo...?
Proyectores: JVC DLA-X7000B, Infocus IN78, Infocus SP4805
Pantalla: Da-Lite Contour Electrol 106" (gan. 1,1)
Fuentes: Chinoppo M9702 V2, HTPC, Server HP Proliant
Receptor AV: Yamaha RX-A3030
Cajas: 2-Keft Q55.2, 1-Keft Cresta C, 2-Keft Cresta 2, 2-B&W 686, 1-Subwoofer Rel QUAKE
Y estas son las hechas para autocal...
Made with high lamp, iris fully open (around 200nits actual peak white):
With these files, I have brightness at +1 and contrast at 0 on the JVC. I have contrast at -1 and brightness at +2 on the Panny.
[EDIT 02-15-17: I now use brightness and contrast both set to zero in the JVC and brightness=+6 and contrast =-1 in the Panny].
02-09-17: 200-1200-1200. This curve clips content above 1200nits. I wouldn't suggest using it.
https://dl.dropboxusercontent.com/u/...-1200-1200.jgd
02-11-17: 200-1200-4000. This curve uses a roll-off from 70% and resolves up to 4000nits. It can be used to play all titles, irrespective of the way they were mastered.
https://dl.dropboxusercontent.com/u/...-1200-4000.jgd
02-12-17: 190-1000-4000. This curve uses a 1000nits target so should be brighter. It resolves up to 4000nits (Universal curve)
https://dl.dropboxusercontent.com/u/...-1000-4000.jgd
02-13-17: 190-1000-1000. This curve is the same as above but without a roll-off. It clips content above 1000nits. You can use this one as a one-for-all curve if you want to priviledge contrast over detail in the highlights, or you can select it manually when you know the title doesn't have any content above 1000nits and want to eek the last performance out of the display. I wouldn't use that curve on its own, as there is quite a lot of titles with content over 1000nits. But it's a great curve to use if yo know there is no content above there.
https://dl.dropboxusercontent.com/u/...-1000-1000.jgd
02-14-17: 200-400-4000. This is a proof of concept curve to show that you can get all the brightness you want (and still resolve the highlights, kind of), but the cost is a very undersaturated picture. I don't recommend to use HDR if you can't target at least 600nits, preferable 800nits. In high lamp, with a peak brightness of 200nits, it means a mulitplier of 3-4 at a minimum. I suggest you reduce contrast to zero in the PJ and set brightness to +1 (assumingHDMI Standard). I had the panny set to contrast -2 for that curve.
https://dl.dropboxusercontent.com/u/...0-400-4000.jgd
The three curves below are discussed in more details here: Official JVC RS600 / RS500 (X950R / X750R - X9000 / X7000) Owners Thread
I now use brightness and contrast both set to zero in the JVC and brightness=+6 and contrast =-1 in the Panny.
02-15-17: 200-1100-4000. As I've decided to use High Lamp, I've made this curve which doesn't bring anything major. It's just the latest I've done for my setup, so feel free to try it.
https://dl.dropboxusercontent.com/u/...-1100-4000.jgd
02-15-17: 200-1100-1100. Here is the same curve, with a sharper roll-off. I've slightly changed my method as I found that clipping hard at 80% leads to unnatural extreme highlights. This method should provide similar results, but it will resolve a bit higher, without clipping that hard. I don't recommend adjusting contrast to clip lower than this curve does for best results.
https://dl.dropboxusercontent.com/u/...-1100-1100.jgd
02-15-17: 200-ST2390-4000. This is an attempt at creating an ST2390 curve (many thanks to Rak306 for providing the targets for this curve as it's not supported in Calman yet). I don't like the results, way too bright and undersaturated in my setup, but others might like it, especially those with some ambient light (it comes out of black much faster) and some with very large screens / low gain screens. It's an alternative to the 200-400-4000 for those in that situation.
https://dl.dropboxusercontent.com/u/...T2390-4000.jgd
Made with low lamp, iris fully open (around 140nits actual peak white):
With these files, I have brightness at +1 and contrast at 0 on the JVC. I have contrast at -2 and brightness at +7 on the Panny.
02-13-17: 140-1100-4000. This curve is a universal curve that can play all titles, irrespective of the way they were mastered, from 1000 to 4000nits. If you don't want any hassle, that the curve to use in low lamp if you have a fairly small screen like I do. The only drawback it a very slight loss of brightness/contrast for titles not having content above 1100nits (not all titles mastered to 4000nits do). The overall brightness of the curve is almost the same.
https://dl.dropboxusercontent.com/u/...-1100-4000.jgd
02-13-17: 140-1100-1100. This curve is the same as above but without a roll-off. It clips content above 1100nits. You can use this one as a one-for-all curve if you want to priviledge contrast over detail in the highlights, or you can select it manually when you know the title doesn't have any content above 1100nits and want to eek the last performance out of the display. Better with a small screen. There are less than 25% of titles (currently) that could have content clipped above 1100nits.
https://dl.dropboxusercontent.com/u/...-1100-1100.jgd
02-13-17: 140-1000-4000. This curve uses a 1000nits target so should be brighter. It resolves up to 4000nits (Universal curve).
https://dl.dropboxusercontent.com/u/...-1000-4000.jgd
02-13-17: 140-1000-1000. This curve is the same as above but without a roll-off. It clips content above 1000nits. You can use this one as a one-for-all curve if you want to priviledge contrast over detail in the highlights, or you can select it manually when you know the title doesn't have any content above 1000nits and want to eek the last performance out of the display. I wouldn't use that curve on its own, as there is quite a lot of titles with content over 1000nits. But it's a great curve to use if you know there is no content above there.
https://dl.dropboxusercontent.com/u/...-1000-1000.jgd
02-14-17: 140-1100-4000 V2. This is a revised version of the curve above of the same name. It should be marginally brighter and more accururate. This is my default curve now. It works fine in low lamp (140nits here) and high lamp (200nits here).
https://dl.dropboxusercontent.com/u/...-4000%20V2.jgd
Todo esto lo digo porque no acabo de pillar si es posible importar las gammas nuevas con el jvcprojector tools...
Estas son las espartanas instrucciones:
https://github.com/arvehj/jvcprojectortools
Y cuando leo esto en avsforum de gente que prueba gammas de las nuevas:JVC Projector Tools
Scripts to send commands 2015 model JVC D-ILA Projectors.
Prerequisites
- python3 installed
- Network connected JVC projector
Tools
jvc_gamma.py
Run to get a menu that lets you generate and load custom gamma curves. Or call load_gamma from the python shell to load other gamma curves.
Parameters for "custom hdr_pq" menu entry
- "max brightness" is the peak brightness your projector can reach.
- "reference white" is how bright you want a 100 nit reference white to be. Lowering this value will make the image darker and increasing it will make the image brighter.
- "soft clip start brightness" is the brightness where the curve will stop following the eotf curve and start a soft clip curve.
- "hard clip brightness" is where the soft clip curve ends.
- "clip end slope" moves the angle of the curve at the end point from horizontal (1) to pointing at the soft clip start point (0).
- "soft clip method" selects between a cubic Bézier curve (0) and quadratic Bézier curve (1).
- "soft clip gamma" selects the gamma to draw the soft clip curve in.
Select toggle white in the menu to match the input level setting in the projector settings you plan to use.
Finally got around to testing Manni's 2-17 JVC-xxx-xxx-4000 curves and pretty much settled on JVC-800-600-4000 as my preferred curve in High Lamp, IRIS -5, and color 0. I was able to lower BR a few clicks too. Plenty of detail and punch watching Pacific Rim for about 20 minutes. My old eyes didn't see any issues really, just nice a very pleasing picture. I like it just slightly more than the older V2 120-1100-4000 which is also very good, but slightly darker with my setup.
I'm going to stick with the JVC-800-600-4000 curve for now and hopefully, Manni or someone will tweak it slightly to eliminate the reported slightly raised black floor or come out with one that's better. The coming Linker will be gravy.
Yesterday that was my preferred gamma curve. But tonight I preferred the Custom PQ Gamma Manni JVC-1000-240-4000 curve from the zip. It looks more natural to me. I was really pleased with how Deadpool looked with it.
Deduzco que se pueden importar ¿pero cómo????
Haz lo que consideres oportuno. Creo que en mis posts anteriores dejo claro el qué y el cómo. En cuanto al "soft rupestre", hace lo que tiene que hacer muy bien.
Saludos
Entiendo que la paciencia se acaba, LTD, perdona.
Revisando tus valiosas aportaciones aquí es donde explicas el funcionamiento de la app. Lo intentaré de nuevo esta noche.
Deduzco entonces que si no hay opción de importar un fichero directamente, si queremos probar las curvas de Manni lo que hay que hacer es copiar manualmente los valores uno por uno ¿es correcto?
Donde dices: "también debe estar seleccionada la gamma de usuario donde quieres cargar el resultado de la ejecución del programa y con la opción interfaz"
¿no te refieres a la opción Importar? (igual que al importar con autocal)
Perrdona tú porque te dije que te echaría las manos que hiciesen falta. A ver King, yo entiendo que a todos (a mí el primero) nos da pereza leernos tochos. Yo, con mayor o menor fortuna, he intentado dar todo mascadito. Y todo lo que preguntas ya lo he explicado. Pero lo voy a repetir.
- Las gammas de Manni se cargan con el autocal. Son gammas para el proyector de Manni y te pueden cuadrar o no. Si quieres que te cuadren mejor tendrás que saber que pico de luz da tu proyector para elegir la que más se adecúe.
- Las gammas de Manni que empiezan con JVC, son gammas generadas con la herramienta rupestre (256 puntos) y después exportadas con el autocal que las capa a 11 puntos. Si las quieres utilizar se cargan como cualquier otra.
- La herramienta rupestre es mucho más ágil que el autocal. Metes los valores y te la carga. Te permite hacer pruebas de manera ágil y rápida.
- Y sí, me he equivocado, quería escribir Importar. Sorry.
Gracias, este era el quid de la cuestión. De si podría importar con auocal las "nuevas" sin cargarme nada.
De momento quiero hacer pruebas rápidas para valorar gammas "basicas" alternativas a la gamma D. Más adelante, con más ganas y tiempo sí que me pondría a mirar lo de generar una propia.
Lo dicho, perdona si pensabas que no hacía caso a tus mensajes, solamente pedía confirmación de si podía probar las custom gammas más recientes (aunque se pierda buena parte por el camino) con seguridad y con autocal.
Simplement decir que la herramienta rupestre no rompe nada. No hay que tenerle miedo. De verdad que es más rápido que meter gammas de manni. Es meter 4 valores y te la carga en el proyector, no te tienes que bajar ningún fichero. En fin tu pispo
Última edición por Anonimo08012019; 20/02/2017 a las 13:57
Vale, vale, también me pondré a ello. Gracias por tu paciencia, de verdad. Da bastante respeto meterle cosas al proyector externamente y ayer lo hice por primera vez. La primera vez de todo siempre cuesta un poco más.
![]()
Hola. Unas imágenes del programa "rupestre" oficialmente bautizado...jeje.
No lo he probado aun en el proyector, pero trasteándolo en el pc, es muy muy simple y super rápido, como comenta LTD. Y hace una gráfica fabulosa. Creo que merece la pena probar este RUPESTRE.
Sldos.
Proyectores: JVC DLA-X7000B, Infocus IN78, Infocus SP4805
Pantalla: Da-Lite Contour Electrol 106" (gan. 1,1)
Fuentes: Chinoppo M9702 V2, HTPC, Server HP Proliant
Receptor AV: Yamaha RX-A3030
Cajas: 2-Keft Q55.2, 1-Keft Cresta C, 2-Keft Cresta 2, 2-B&W 686, 1-Subwoofer Rel QUAKE
esto es lo que funciona seguro, lo demas no lo se:
con esto actualize mi X7000 sin problemas con otros adaptadores no hubo forma.
https://www.amazon.es/gp/product/B00...?ie=UTF8&psc=1
https://www.amazon.es/gp/product/B00...?ie=UTF8&psc=1
Proyector jvc x7000 y lente Prismasonic HF1500
configuracion Atmos 7.2.4
Altavoces KCS trio frontal C215F
Subwoofers KCS C118 x 2 svs pc-13 ultra y jbl gto 1514
Efecto KCS SR10 x10
Pantalla SMX transparent audio 178" 2:37:1 curvada
procesador Marantz AV7702mkII
Etapas Crown XLS 1000, XLS 2500
The t.amp D4-500, behringer inuke6000dsp
Yamaha P7000s, Yamaha px3 y the t.amp E4 250
Lector: panasonic UB900 UHD
MC: Himedia 10 pro
cable hdmi Velox
Ok con el rupestre soft, probaré a generar mi propia gamma con los valores cercanos a los propuestos.
Gracias de verdad por el esfuerzo en explicar las cosas.
Proyectores: JVC DLA-X7000B, Infocus IN78, Infocus SP4805
Pantalla: Da-Lite Contour Electrol 106" (gan. 1,1)
Fuentes: Chinoppo M9702 V2, HTPC, Server HP Proliant
Receptor AV: Yamaha RX-A3030
Cajas: 2-Keft Q55.2, 1-Keft Cresta C, 2-Keft Cresta 2, 2-B&W 686, 1-Subwoofer Rel QUAKE
.
Última edición por jlakira; 11/07/2018 a las 14:59
Estoy ahora mismo con el rupestre tool y ya llevo 3 gammas creadas con el soft clip a 600, 800 y 1000 respectivamente.
Se hace duro el temita probando y probando, pero estoy en ello.
De momento consigo los mejores resultados con el input en standard en el JVC y los siguientes valores el el rupestre soft:
max brigthness 175
reference white 15.91
soft clip 600
hard clip 4000
1
1
1
y subiendo un poco el brillo y contraste en el jvc.
Negros culo de orco y buenos colores en los siete magníficos. Sigo probando.
Tengo la intención de vivir para siempre, o morir en el intento.